MediumNLPnlp-394

METEOR Weighted F Mean

Problem

The METEOR-lite score combines precision and recall as a weighted harmonic mean: P R / (alpha P + (1 - alpha) R).

Given alpha, precision, and recall, return the score; return 0.0 when the denominator is zero.

Hint

Recall is weighted more heavily for alpha near one.

case 1

in: [0.9,0.8,0.6]

expected: 0.6153846153846153

case 2

in: [0.5,0.5,0.5]

expected: 0.5

case 3

in: [0.9,0,0]

expected: 0

Starter codePython

def meteor_weighted_f_mean(alpha, precision, recall):
    # Your code here
    pass

3 test cases · run your code against them in the editor